Method and apparatus for recommending multimedia resource

1. A method performed in a server for recommending a multimedia resource, comprising:

acquiring user viewing data of a to-be-recommended multimedia resource in a playing trial run, wherein the user viewing data includes at least user audio and video data and user video depth data;

analyzing the user viewing data to obtain user action data corresponding to the to-be-recommended multimedia resource;

calculating a first user preference score of the to-be-recommended multimedia resource based on the user action data;

acquiring a plurality of second user preference scores of a plurality of recommended multimedia resources; and

recommending the to-be-recommended multimedia resource based on the first user preference score and the second user preference scores,

wherein the calculating the first user preference score further includes:

determining at least one user action from the user action data;

inquiring a pre-stored correlation between user actions and user preference values, to obtain a user preference value corresponding to the at least one user action;

determining a weight corresponding to the at least one user action based on a ratio of a duration of the at least one user action to a playing time length of the to-be-recommended multimedia resource; and

calculating the first user preference score based on the user preference value and the weight, and

wherein before inquiring the pre-stored correlation between the user actions and the user preference values, the method further comprises:

pre-determining a user preference value for each of the user actions; and

storing the correlation between the user actions and the user preference values,

wherein the recommending of the to-be-recommended multimedia resource comprises:

among the plurality of second user preference scores, determining a third user preference score which is similar to the first user preference score;

determining a user to receive the to-be-recommend multimedia resource based on the third user preference score; and

sending the to-be-recommended multimedia resource to a playing device of the receiving user,

wherein the second user preference scores are calculated based on data collected by a plurality of sensors mounted on or in vicinity of a plurality of playing devices corresponding to the recommended multimedia resources.

2. The method according to claim 1 , wherein the determining of the user to receive the to-be-recommended multimedia resource based on the third user preference score comprises:

determining that the third user preference score is greater than a pre-determined threshold; and

determining a user corresponding to the third user preference score which is greater than the pre-determined threshold as the user to receive the to-be-recommended multimedia resource.

3. The method according to claim 1 , wherein the user viewing data is first user viewing data and, after sending the to-be-recommended multimedia resource to the playing device of the receiving user, the method further comprises:

acquiring second user viewing data collected by a sensor corresponding to the to-be-recommended multimedia resource;

analyzing the second user viewing data to obtain a plurality pieces of second user action data corresponding to the to-be-recommended multimedia resource;

based on each piece of the second user action data, updating the first user preference score of the to-be-recommended multimedia resource; and

determining effectiveness of the to-be-recommended multimedia resource based on the updated first user preference score.

4. The method according to claim 1 , wherein the user viewing data is collected by a plurality of sensors which are mounted on or in vicinity of a plurality of playing devices in the playing trial run.

5. A server for recommending a multimedia resource, comprising:

a processor; and

a memory for storing instructions executable by the processor;

wherein the processor is configured to:

acquire user viewing data of a to-be-recommended multimedia resource in a playing trial run, wherein the user viewing data includes at least user audio and video data and user video depth data;

analyze the user viewing data to obtain user action data corresponding to the to-be-recommended multimedia resource;

calculate a first user preference score of the to-be-recommended multimedia resource based on the user action data;

acquire a plurality of second user preference scores of a plurality of recommended multimedia resources; and

recommend the to-be-recommended multimedia resource based on the first user preference score and the second user preference scores,

wherein, in calculating the first user preference score, the processor is further configured to:

determine at least one user action from the user action data;

inquire a pre-stored correlation between user actions and user preference values, to obtain a user preference value corresponding to the at least one user action;

determine a weight corresponding to the at least one user action based on a ratio of a duration of the at least one user action to a playing time length of the to-be-recommended multimedia resource; and

calculate the first user preference score based on the user preference value and the weight,

wherein before inquiring the pre-stored correlation between the user actions and the user preference values, the processor is further configured to:

pre-determine a user preference value for each of the user actions; and

store the correlation between the user actions and the user preference values,

wherein the processor is further configured to:

among the plurality of second user preference scores, determine a third user preference score which is similar to the first user preference score;

determine a user to receive the to-be-recommended multimedia resource based on the third user preference score; and

send the to-be-recommended multimedia resource to a playing device of the receiving user,

wherein the second user preference scores are calculated based on data collected by a plurality of sensors which are mounted on or in vicinity of a plurality of playing devices corresponding to the recommended multimedia resources.

6. The server according to claim 5 , wherein the processor is further configured to:

determine that the third user preference score is greater than a pre-determined threshold; and

determine a user corresponding to the third user preference score which is greater than the pre-determined threshold as the user to receive the to-be-recommended multimedia resource.

7. The server according to claim 5 , wherein the user viewing data is first user viewing data, and the processor is further configured to:

acquire second user viewing data collected by a sensor corresponding to the to-be-recommended multimedia resource;

analyze the second user viewing data to obtain a plurality pieces of second user action data corresponding to the to-be-recommended multimedia resource;

based on each piece of the second user action data, update the first user preference score of the to-be-recommended multimedia resource; and

determine effectiveness of the to-be-recommended multimedia resource based on the updated first user preference score.

8. The server according to claim 5 , wherein the user viewing data is collected by a plurality of sensors which are mounted on or in vicinity of a plurality of playing devices in the playing trial run.

9. A non-transitory storage medium having stored therein instructions that, when executed by one or more processors of a server for recommending a multimedia resource, cause the server to perform:

acquiring user viewing data of a to-be-recommended multimedia resource in a trial playing run, wherein the user viewing data includes at least user audio and video data and user video depth data;

analyzing the user viewing data to obtain user action data corresponding to the to-be-recommended multimedia resource;

calculating a first user preference score of the to-be-recommended multimedia resource based on the user action data;

acquiring a plurality of second user preference scores of a plurality of recommended multimedia resources; and

recommending the to-be-recommended multimedia resource based on the first user preference score and the second user preference scores,

wherein, in calculating the first user preference score, the instructions are executed to cause the server to further perform:

determining at least one user action from the user action data;

inquiring a pre-stored correlation between user actions and user preference values, to obtain a user preference value corresponding to the at least one user action;

determining a weight corresponding to the at least one user action based on a ratio of a duration of the at least one user action to a playing time length of the to-be-recommended multimedia resource; and

calculating the first user preference score based on the user preference value and the weight,

wherein before inquiring the pre-stored correlation between the user actions and the user preference values, the instructions are executed to cause the server to further perform:

pre-determining a user preference value for each of the user actions; and

storing the correlation between the user actions and the user preference values,

wherein the instructions are executed to cause the server to further perform:

among the plurality of second user preference scores, determining a third user preference score which is similar to the first user preference score;

determining a user to receive the to-be-recommended multimedia resource based on the third user preference score; and

sending the to-be-recommended multimedia resource to a playing device of the receiving user,

wherein the second user preference scores are calculated based on data collected by a plurality of sensors which are mounted on or in vicinity of a plurality of playing devices corresponding to the recommended multimedia resources.